Reagent Control of Geometric Selectivity and Enantiotopic Group

X Royal Danish School of Pharmacy. ... (v) Tanaka, K.;. Otsubo, K.; Fuji, K. Tetrahedron Lett. 1996, 37, 3735-3738. (x) ... (aa) Mendlik, M. T.; Cotta...
0 downloads 0 Views 155KB Size
Download PDF
8284

J. Org. Chem. 1998, 63, 8284-8294

Reagent Control of Geometric Selectivity and Enantiotopic Group Preference in Asymmetric Horner-Wadsworth-Emmons Reactions with meso-Dialdehydes† Joshua S. Tullis,‡,§,| Lauri Vares,‡,§ Nina Kann,§,⊥ Per-Ola Norrby,X and Tobias Rein*,‡,§ Department of Organic Chemistry, Technical University of Denmark, DK-2800 Lyngby, Denmark, Department of Organic Chemistry, Royal Institute of Technology, S-100 44 Stockholm, Sweden, Department of Chemistry and Biochemistry, University of Notre Dame, Notre Dame, Indiana 46556, and Department of Medicinal Chemistry, Royal Danish School of Pharmacy, Universitetsparken 2, DK-2100 Copenhagen, Denmark Received June 9, 1998

Results from asymmetric Horner-Wadsworth-Emmons reactions between chiral phosphonate reagents 3a-d, which contain (1R,2S,5R)-8-phenylmenthol as a chiral auxiliary, and mesodialdehydes 6 and 14 are presented. It was found that both the geometric selectivities and the levels of asymmetric induction depended on the structure of the phosphonate (i.e., the alkyl group R1 in the phosphoryl unit) and to a certain extent also on the reaction conditions. Furthermore, the nature of the protecting group used on the R-oxygen substituent in dialdehydes 14 influenced the outcome somewhat. By an appropriate choice of reagent and conditions, either (E)- or (Z)monoaddition products could be obtained geometrically pure and with good to excellent diastereoselectivities, in synthetically useful yields. Analyses of the absolute configurations of the products showed that the (E)-selective reagents (3a-c) and the (Z)-selective phosphonate 3d reacted at opposite enantiotopic carbonyl groups in the substrates. A mechanistic model which accounts for the products formed is presented. Introduction Selective reaction of only one of two enantiotopic groups in a bifunctional substrate is a powerful strategy1 for asymmetric synthesis, as witnessed by the increasing attention such processes have received in recent years; a variety of examples, using both enzymatic2 and nonenzymatic3 reactions, has been reported. Based on this concept, reaction types in which no additional sp3 stereocenter is created at any of the bond-forming sites can also be used for asymmetric synthesis, since asymmetric induction is achieved by “desymmetrization” of the substrate. One such class of reactions is asymmetric Wittig-type reactions, an area which in recent years has been studied by a number of research groups,4 with several examples of highly selective transformations being reported. In the vast majority of asymmetric * To whom correspondence should be addressed. E-mail address: [email protected]. † This paper is gratefully dedicated to our colleague and friend Professor Paul Helquist, University of Notre Dame. ‡ Technical University of Denmark. § Royal Institute of Technology. | University of Notre Dame. ⊥ Present address: Department of Organic Chemistry, Chalmers University of Technology, S-412 96 Go¨teborg, Sweden. X Royal Danish School of Pharmacy. (1) For reviews on the use of bifunctional substrates in twodirectional synthesis, see: (a) Poss, C. S.; Schreiber, S. L. Acc. Chem. Res. 1994, 27, 9-17, and references therein. (b) Magnuson, S. R. Tetrahedron 1995, 51, 2167-2213. (c) Maier, M. In Organic Synthesis Highlights II; Waldmann, H., Ed.; VCH: Weinheim, 1995; pp 203222. (2) For an excellent recent review, see: Schoffers, E.; Golebiowski, A.; Johnson, C. R. Tetrahedron 1996, 52, 3769-3826. (3) (a) Eder, U.; Sauer, G.; Wiechert, R. Angew. Chem. 1971, 83, 492-493. (b) Hajos, Z. G.; Parrish, D. R. J. Org. Chem. 1974, 39, 16151621. For reviews, see: (c) Ward, R. S. Chem. Soc. Rev. 1990, 19, 1-19. (d) Gais, H.-J. In Houben-Weyl, Stereoselective Synthesis; Helmchen, G., Hoffman, R. W., Mulzer, J., Schaumann, E., Eds.; Thieme: Stuttgart, 1996; Vol. 1, pp 589-643.

Wittig-type reactions studied so far, however, monocarbonyl compounds have been utilized as substrates; only in a few cases4f,n,q,bb have prochiral diketones been employed, and prior to our first report4m the possibility (4) For a review, see: (a) Rein, T.; Reiser, O. Acta Chem. Scand. 1996, 50, 369-379. See also: (b) Li, A.-H.; Dai, W.-M.; Aggarwal, V. K. Chem. Rev. 1997, 97, 2341-2372. (c) Bennani, Y. L.; Hanessian, S. Chem. Rev. 1997, 97, 3161-3195. (d) Nicolaou, K. C.; Ha¨rter, M. W.; Gunzner, J. L.; Nadin, A. Liebigs Ann./Recueil 1997, 1283-1302. For selected examples, see: (e) To¨mo¨sko¨zi, I.; Janszo´, G. Chem. Ind. (London) 1962, 2085-2086. (f) Trost, B. M.; Curran, D. P. Tetrahedron Lett. 1981, 22, 4929-4932. (g) Hanessian, S.; Delorme, D.; Beaudoin, S.; Leblanc, Y. J. Am. Chem. Soc. 1984, 106, 5754-5756. (h) Gais, H.J.; Schmiedl, G.; Ball, W. A.; Bund, J.; Hellmann, G.; Erdelmeier, I. Tetrahedron Lett. 1988, 29, 1773-1774. (i) Rehwinkel, H.; Skupsch, J.; Vorbru¨ggen, H. Tetrahedron Lett. 1988, 29, 1775-1776. (j) Hanessian, S.; Beaudoin, S. Tetrahedron Lett. 1992, 33, 7655-7658. (k) Denmark, S. E.; Chen, C.-T. J. Am. Chem. Soc. 1992, 114, 1067410676. (l) Narasaka, K.; Hidai, E.; Hayashi, Y.; Gras, J.-L. J. Chem. Soc., Chem. Commun. 1993, 102-104. (m) Kann, N.; Rein, T. J. Org. Chem. 1993, 58, 3802-3804. (n) Tanaka, K.; Ohta, Y.; Fuji, K.; Taga, T. Tetrahedron Lett. 1993, 34, 4071-4074. (o) Rein, T.; Kann, N.; Kreuder, R.; Gangloff, B.; Reiser, O. Angew. Chem. 1994, 106, 597599; Angew. Chem., Int., Ed. Engl. 1994, 33, 556-558. (p) Furuta, T.; Iwamura, M. J. Chem. Soc., Chem. Commun. 1994, 2167-2168. (q) Mandai, T.; Kaihara, Y.; Tsuji, J. J. Org. Chem. 1994, 59, 5847-5849. (r) Denmark, S.; Rivera, I. J. Org. Chem. 1994, 59, 6887-6889. (s) Rein, T.; Kreuder, R.; von Zezschwitz, P.; Wulff, C.; Reiser, O. Angew. Chem. 1995, 107, 1099-1102; Angew. Chem., Int. Ed. Engl. 1995, 34, 1023-1025. (t) Rein, T.; Anvelt, J.; Soone, A.; Kreuder, R.; Wulff, C.; Reiser, O. Tetrahedron Lett. 1995, 36, 2303-2306. (u) Abiko, A.; Masamune, S. Tetrahedron Lett. 1996, 37, 1077-1080. (v) Tanaka, K.; Otsubo, K.; Fuji, K. Tetrahedron Lett. 1996, 37, 3735-3738. (x) Kumamoto, T.; Koga, K. Chem. Pharm. Bull. 1997, 45, 753-755. (y) Gais, H.-J.; Schmiedl, G.; Ossenkamp, R. K. L. Liebigs Ann./Recueil 1997, 2419-2431. (z) Dai, W.-M.; Wu, J.; Huang, X. Tetrahedron: Asymmetry 1997, 8, 1979-1982. (aa) Mendlik, M. T.; Cottard, M.; Rein, T.; Helquist, P. Tetrahedron Lett. 1997, 38, 6375-6378. (bb) Tanaka, K.; Watanabe, T.; Ohta, Y.; Fuji, K. Tetrahedron Lett. 1997, 38, 89438946. (cc) Kreuder, R.; Rein, T.; Reiser, O. Tetrahedron Lett. 1997, 38, 9035-9038. (dd) Mizuno, M.; Fujii, K.; Tomioka, K. Angew. Chem. 1998, 110, 525-527; Angew. Chem., Int. Ed. 1998, 37, 515-517. (ee) Arai, S.; Hamaguchi, S.; Shioiri, T. Tetrahedron Lett. 1998, 39, 29973000.

10.1021/jo981102z CCC: $15.00 © 1998 American Chemical Society Published on Web 10/27/1998

Asymmetric Horner-Wadsworth-Emmons Reactions

of using dialdehyde substrates5 had not been investigated. From the viewpoint of synthetic efficiency, this alternative is appealing since the reaction product directly can take part in an additional chain-extending reaction involving the unreacted aldehyde, as exemplified in Scheme 1. Thus, relatively complex chiral structures should be quickly accessible from nonchiral precursors. In this paper, we report results from an extension of our earlier6 studies of asymmetric Horner-WadsworthEmmons (HWE) reactions7 with dialdehydes and demonstrate that complementary product selectivities are possible by an appropriate choice of reagent: by slight structural variation in the chiral phosphonate, either (E)or (Z)-alkenes can be obtained as products, with both high geometric selectivity and good to excellent levels of asymmetric induction. Furthermore, the (E)- and (Z)-selective reagents are complementary also in the sense that they react with opposite enantiotopic group preference.

J. Org. Chem., Vol. 63, No. 23, 1998 8285 Scheme 1

Results Choice and Preparation of Chiral Phosphonate Reagents. For our initial studies, we chose to examine reactions between suitable model dialdehyde substrates and the chiral phosphonates 3a-d, derived from (1R,2S,5R)-8-phenylmenthol (eq 1). Reagent 3a has earlier been shown to give useful levels of diastereose-

lectivity in reactions with a prochiral monoketone4h,y and with some structurally related chiral ketones.4h,i Furthermore, we felt that incorporation of a chiral auxiliary in the anion-stabilizing functionality rather than in the phosphorus-based functional group could give larger synthetic versatility, since choice of different alkyl groups R1 in the reagent then might enable control of the alkene (5) For examples of the use of meso-dialdehydes as substrates for other types of asymmetric transformations, see: (a) Yamazaki, Y.; Hosono, K. Tetrahedron Lett. 1988, 29, 5769-5770. (b) Roush, W. R.; Park, J. C. Tetrahedron Lett. 1990, 31, 4707-4710. (c) Wang, Z.; Descheˆnes, D. J. Am. Chem. Soc. 1992, 114, 1090-1091. (d) Ward, D. E.; Liu, Y.; Rhee, C. K. Synlett 1993, 561-563. (e) Ward, D. E.; Liu, Y. D.; Rhee, C. K. Can. J. Chem. 1994, 72, 1429-1446. (f) Roush, W. R.; Wada, C. K. J. Am. Chem. Soc. 1994, 116, 2151-2152. (g) Oppolzer, W.; De Brabander, J.; Walther, E.; Bernardinelli, G. Tetrahedron Lett. 1995, 36, 4413-4416. (h) De Brabander, J.; Oppolzer, W. Tetrahedron 1997, 53, 9169-9202. (i) Oppolzer, W.; Walther, E.; Balado, C. P.; De Brabander, J. Tetrahedron Lett. 1997, 38, 809-812. (6) Part of this work has been described in an earlier communication; see ref 4m.

geometry in the product. This expectation was realized in practice. A procedure for preparation of the chiral phosphonates 3a and 3c by transesterification of the achiral precursors 1a and 1c with 8-phenylmenthol (2) has previously been reported by Takano and co-workers.8 In the same manner, we prepared4o reagents 3b and 3d in good yields from the corresponding known phosphonates 1b and 1d.9 Choice and Preparation of Model Substrates. As our model dialdehyde substrates, we chose compounds 610 and 14. The motivation for this choice was 2-fold: (i) reactions with these substrates would give information on the influence of different types of R-substituents (CH3 or RO) in the dialdehyde on the reaction selectivity; (ii) the expected chiral monoaddition products are potentially useful synthetic intermediates, as they correspond to partial structures of a number of natural products of biomedical interest. Thus, the products expected from dialdehyde 6 match subunits of several strongly cytotoxic macrolides,11 whereas products derived from dialdehydes of type 14 can be envisioned as building blocks for polyene/polyol macrolide antibiotics.12 (7) (a) Horner, L.; Hoffman, H.; Wippel, H. G. Chem. Ber. 1958, 91, 61-63. (b) Horner, L.; Hoffman, H.; Wippel, H. G.; Klahre, G. Chem. Ber. 1959, 92, 2499-2505. (c) Wadsworth, W. S.; Emmons, W. D. J. Am. Chem. Soc. 1961, 83, 1733-1738. For reviews discussing general aspects of the HWE reaction, see: (d) Boutagy, J.; Thomas, R. Chem. Rev. 1974, 74, 87-99. (e) Wadsworth, W. S., Jr. Org. React. 1977, 25, 73-253. (f) Walker, B. J. In Organophosphorus Reagents in Organic Synthesis; Cadogan, J. I. G., Ed.; Academic Press: New York, 1979; pp 155-205. (g) Maryanoff, B. E.; Reitz, A. B. Chem. Rev. 1989, 89, 863-927. (h) Kelly, S. E. In Comprehensive Organic Synthesis; Trost, B. M., Fleming, I., Eds.; Pergamon Press: Oxford, 1991; Vol. 1, pp 730817. See also ref 4d. (8) Hatakeyama, S.; Satoh, K.; Sakurai, K.; Takano, S. Tetrahedron Lett. 1987, 28, 2713-2716. In our hands, stoichiometric amounts of DMAP were in some cases required for the reactions to give good yields. (9) Compound 1b is commercially available. Phosphonate 1d was prepared according to the procedure of Still: Still, W. C.; Gennari, C. Tetrahedron Lett. 1983, 24, 4405-4408. (10) After our initial report on asymmetric HWE reactions with 6, this dialdehyde has also been used as substrate in group-selective asymmetric aldol reactions; see refs 5g and 5h. (11) Norcross, R. D.; Paterson, I. Chem. Rev. 1995, 95, 2041-2114. (12) Rychnovsky, S. D. Chem. Rev. 1995, 95, 2021-2040.

8286 J. Org. Chem., Vol. 63, No. 23, 1998 Scheme 2

Tullis et al. Scheme 4

Scheme 3

Diol 5, the immediate precursor of dialdehyde 6, was obtained by double hydroboration of diene 4,13 as described by Harada and co-workers.14 Compound 6 could be prepared in pure form by Swern oxidation15 of 5 if water was carefully excluded during the workup (Scheme 2). In the presence of water, the cyclic hydrate 7 was isolated as the major product,16 and it proved difficult to regenerate 6 once the hydrate had been formed. Dialdehyde 6 is stable in solution for some time, if stored protected from water; we recommend, however, that it is freshly prepared for use. The other model substrates, dialdehydes 14, were synthesized in five steps from 6-(benzyloxy)-1,3-cycloheptadiene (9) which, in turn, is accessible from 1,3,5cycloheptatriene17 (Scheme 3). The relative stereochemistry of the three stereocenters in 14 was controlled by a (13) For experimental details regarding the preparation of 4 and 5, see Supporting Information. (14) (a) Harada, T.; Matsuda, Y.; Uchimura, J.; Oku, A. J. Chem. Soc., Chem. Commun. 1990, 21-22. (b) Harada, T.; Inoue, A.; Wada, I.; Uchimura, J.; Tanaka, S.; Oku, A. J. Am. Chem. Soc. 1993, 115, 7665-7674. For an enzyme-catalyzed desymmetrization of 5 by groupselective esterification, see: Chenevert, R.; Courchesne, G. Tetrahedron: Asymmetry 1995, 6, 2093-2096. (15) Mancuso, A.; Huang, S.-L.; Swern, D. J. Org. Chem. 1978, 43, 2480-2482. (16) Harada, T.; Kagamihara, Y.; Tanaka, S.; Sakamoto, K.; Oku, A. J. Org. Chem. 1992, 57, 1637-1639.

palladium-catalyzed cis-diacetoxylation18 of diene 9 to give diacetate 10.19 To gain some insight into the influence of different oxygen-protecting groups R in 14 on the outcome of the asymmetric HWE reactions, the tert-butyldiphenylsilyl-protected substrate 14a, and the pivaloyl-protected 14b were then prepared from 10 by straightforward transformations.20 Diols 13a and 13b were both formed as diastereomeric mixtures, but upon treatment with periodic acid, both diastereomers of each compound were cleanly converted to the desired dialdehyde. Dialdehyde 14a is quite robust and stable enough to be purified by chromatography without detectable epimerization. Compound 14b, however, is more sensitive, and is best prepared fresh immediately prior to use. Asymmetric HWE Reactions. To investigate the influence of reactant stoichiometry and reaction temperature, we first studied reactions between dimethylphosphonate 3a and dialdehydes 6 and 14a (Scheme 4, Table 1). Conditions which were expected to favor kinetic control in the initial addition step, by increasing the relative rate of the subsequent elimination, were chosen [potassium hexamethyldisilazide (KHMDS) as base in combination with 18-crown-6, low temperature].21 In these initial experiments, the crude reaction product was reduced with NaBH4 to facilitate separation of the products from both bis-addition products and remaining unreacted dialdehyde. Somewhat to our surprise, essentially only (E)-products were obtained from both substrates, although nonasymmetric HWE reactions with dimethyl phosphonates often give (Z)-products under similar reaction conditions. These initial experiments demonstrated that (E)-alkenes 1522 and 1722 could be obtained in reasonable yields and (17) Schink, H. E.; Petterson, H.; Ba¨ckvall, J.-E. J. Org. Chem. 1991, 56, 2769-2774. (18) Ba¨ckvall, J.-E.; Bystro¨m, S. E.; Nordberg, R. E. J. Org. Chem. 1984, 49, 4619-4631.

Asymmetric Horner-Wadsworth-Emmons Reactions Table 1.

entry 1 2 3 4 5 6 7 8

J. Org. Chem., Vol. 63, No. 23, 1998 8287 Scheme 5

Reactions of Phosphonate 3a with Dialdehydes 6 and 14aa

substrate temp yieldb diaster yield of bis(equiv) (°C) product (%) ratioc addition (%) 6 (2.0) 6 (2.1) 6 (1.2) 6 (1.3) 14a (2.0) 14a (2.0) 14a (1.2) 14a (1.2)

-78 -100 -78 -100 -78 -100 -78 -100

15 15 15 15 17 17 17 17

88 87 36 77 68 50 60d 49

87:13 90:10 97:3 91:9 87:13 89:11 88:12 91:9

5 48 6 e -

a General reaction conditions: 1.1 equiv of phosphonate, 1.0 equiv of KHMDS, 5 equiv of 18-crown-6, ca. 0.02 M in THF, 2.515 h. b Isolated yield of product judged as g95% pure by NMR and TLC, unless otherwise noted. c Ratio in isolated product; ratios in crude products were within (2% of these values. Entries 1-4: ratio 15:16; entries 5-8: ratio 17:18. Geometric ratios were g98: 2. d Another product, tentatively assigned as being the product of silyl group migration from the secondary to the primary hydroxyl, was isolated in 18% yield. e Although small amounts of bis-addition product were detected in the crude product, none were isolated after chromatography.

diastereoselectivities even when close to equimolar amounts of substrate and reagent were used (compare entries 2 and 4, 5 and 7). In addition, the reaction temperature turned out to be an important parameter, as expected. With the more reactive substrate 6, reaction at -78 °C gave considerable amounts of bis-addition product (entry 3); this was largely suppressed by performing the reaction at -100 °C (entry 4). On the other hand, the slightly lower yields obtained from dialdehyde 14a at the lower reaction temperature presumably resulted from incomplete conversion due to this substrate being more sterically hindered (compare entries 6 and 8 with 5 and 7). We then embarked on a more extensive study of reactions involving dialdehydes 6, 14a, and 14b. Depending on the synthetic objective, it is desirable that either (E)- or (Z)-alkenes can be prepared with high selectivity in the asymmetric HWE reactions. For this reason, we investigated the utility of reagents 3a-d containing different alkyl groups R1 in the phosphoryl unit. It is known from standard nonasymmetric HWE reactions7,21 that the alkene geometry of the product often can be controlled by an appropriate choice of R1 and reaction conditions. The results obtained in reactions with 6 (Scheme 5, Table 2) demonstrated that selective preparation of either (E)- or (Z)-products was indeed possible. Phosphonates 3a-c all gave essentially exclusive formation of (E)-products. However, the structure of (19) Diacetate 10 was assigned as meso-(3R,5s,7S)-5-(benzyloxy)3,7-diacetoxycyclohept-1-ene on the basis of 1H NMR data, by comparison with the known18 compound meso-(3R,5s,7S)-5-methoxy-3,7diacetoxycyclohept-1-ene. As evidenced by NMR on the crude product, the Pd-catalyzed diacetoxylation was not completely stereoselective. However, the other product isomers could be cleanly removed by use of the MPLC system described by Baeckstro¨m: Baeckstro¨m, P.; Stridh, K.; Li, L.; Norin, T. Acta Chem. Scand., Ser. B 1987, 41, 442-447. (20) Dihydroxylation of diacetate 10 gave a complex mixture of regioisomeric products, presumably due to facile acyl group migration in the initially formed diol. In contrast, dihydroxylation of bis-pivalate 12b proceeded more cleanly. Upon chromatographic purification, acyl migration sometimes occurred also with diols 13b; however, if the crude diastereomeric mixture of diols 13b was directly cleaved with H5IO6, excellent yields of isomerically pure dialdehyde 14b were reproducibly obtained. Oxidative cleavage of the alkenes 12 by ozonolysis was also attempted, but in our hands the two-step osmylation/periodic acid protocol gave much cleaner products. (21) Thompson, S. K.; Heathcock, C. H. J. Org. Chem. 1990, 55, 3386-3388, and references therein.

Table 2.

Reactions of Phosphonates 3a-d with Dialdehyde 6a

entry phosphonate 1 2 3 4 5 6 7

3a 3b 3c 3d 3d 3d 3d

temp (°C)

product

yieldb (%)

diaster ratioc

yield of bisaddition (%)

-100 -90 -90 -100 -90 -78 0h

19 19 19 21 21 21 21

53 25f 2f 74 83 76 38i

95:5d ca. 95:5d ca. 97:3d g98:2g g98:2 g98:2 g98:2

e 38f 12f e 36i

a General reaction conditions: 1.1 equiv of dialdehyde, 1.1 equiv of phosphonate, 1.0 equiv of KHMDS, 5 equiv of 18-crown-6, ca. 0.02 M in THF, 6-22 h. b Isolated yield of product judged as g95% pure by NMR and TLC, unless otherwise noted. c Ratio in isolated product; ratios in crude products were within (2% of these values. Entries 1-3: ratio 19:20; entries 4-7: ratio 21:22. Geometric ratios were g98:2, unless stated otherwise. d Small amounts of another isomer, assigned as an epimer, were also present in the isolated product: entry 1, 3%; entry 2, 9%; entry 3, 4% (see Supporting Information). e Not determined. f (E)-Mono- and (E,E)bis-addition products were obtained in the same fraction after chromatography. g The isolated product also contained 6% of another isomer, assigned as an epimer; this was not present in the crude, but was formed during chromatography (see Supporting Information). h Reaction time 2 h. i After chromatography, the (Z)monoaddition product 21 (38% yield) and the (E,Z)-bis-addition product 28 (25% yield; see Chart 1) were obtained in the same fraction. In separate fractions, the (Z,Z)- and (E,E)-bis-addition products were isolated, in 9% and 2% yield, respectively.

R1 did influence the yield of monoaddition product strongly: the dimethylphosphonate 3a was clearly superior to the diethyl and diisopropyl analogues. Use of 3a in combination with KHMDS/18-crown-6 gave the (E)monoaddition product 1922,23 in 53% yield with a diastereomer ratio of 95:5 (entry 1). The (Z)-selective reagent 3d proved to be even more efficient than the (E)-selective reagents 3a-c in its reactions with 6. Together with KHMDS/18-crown-6 as (22) For details regarding how structure assignments for the HWE product isomers have been made, and how isomer ratios have been determined, see Supporting Information.

8288 J. Org. Chem., Vol. 63, No. 23, 1998 Scheme 6

base, reagent 3d gave the (Z)-product 2122,23 in high yield, with excellent geometric selectivity and as a single detectable diastereomer (entry 5). The temperature dependence of this reaction was investigated to some extent: reactions at -78 °C and lower afforded good yields of monoaddition product, but a reaction performed at 0 °C gave a much reduced yield due to formation of substantial amounts of bisaddition products. The reactions with the R-oxygenated substrates 14a and 14b (Scheme 6, Table 3) followed similar general trends as the reactions with 6, although there are some differences worth noting. Reagents 3a-c all gave high (E)-selectivities in reactions with both 14a and 14b. The diastereoselectivities observed for the (E)-product 23a22,23 were similar for all three reagents (entries 1-3); although the amount of unreacted dialdehyde was not determined, the slightly lower yield observed in the reaction with the diisopropyl phosphonate 3c may well be due to incomplete conversion. On the other hand, in reactions with the more reactive, pivaloyl-protected substrate 14b, the dimethyl phosphonate 3a performed poorly (entry 7), whereas both 3b and 3c gave 23b22,23 in much better yields and diastereoselectivities (entries 8 and 9). Thanks to the fact that the diastereomers of 23a could be separated (23) In general, the (E)- and (Z)-monoaddition products could be separated by flash chromatography. It also proved possible to separate the (E)-diastereomers 23a and 24a (Scheme 6, vide infra) by chromatography, if Amicon silica (see Experimental Section) was used; however, we have not yet found conditions which enable complete separation of other diastereomeric products (i.e., 19/20, 23b/24b, 25a/ 26a). Some of the HWE products showed tendencies to undergo slight epimerization during chromatography, but this could be suppressed by use of appropriate conditions: compounds 19 and 21 could be chromatographed on Merck silica (see Experimental Section) if the silica was deactivated by elution with EtOAc or EtOAc/MeOH prior to chromatography; compounds 23b and 25b could be chromatographed on Amicon silica without detectable epimerization, whereas use of the Merck silica caused some epimerization. The silyl-protected products 23a and 25a did not undergo observable epimerization on either type of silica.

Tullis et al.

under optimized chromatography conditions, diastereomerically pure 23a was obtained in a synthetically useful yield, even though the diastereoselectivity in the crude product was slightly lower than for 23b. The bis(trifluoroethyl)phosphonate 3d provided access to the (Z)-products 25a22,23 and 25b22,23 with excellent levels of asymmetric induction and in good yields. The reactions with aldehyde 14a proved to be special cases: both KHMDS and NaHMDS alone gave better results than our “standard” base system KHMDS/18-crown-6 (entries 4-6). The reaction between 3d and 14b, however, gave best selectivities under the standard conditions (entry 10). Determination of Absolute Configurations. The assignments of absolute configurations for compounds 15 and 19 are based on 1H NMR analyses of both diastereomers of the Mosher ester derivative 27 (Chart 1), according to the method introduced by Mosher and Dale and extended by Kakisawa and co-workers.24 Compound 21 is assigned the indicated absolute configuration based on a correlation with 19: both 19 and 21 were converted to the same mono-(Z)/mono-(E)-diastereomer 28 by reaction with 3d and 3a, respectively. On the basis of NMR analysis, the products from these two reactions were identical, which, in turn, implies that the monoaddition products 19 and 21 must have been formed by reaction at enantiotopic carbonyl groups in 6. In contrast, if the (E)- and (Z)-monoaddition products were formed via reaction at the same carbonyl group, further reaction of the (Z)-product with 3a would have given the diastereomeric mono-(Z)/mono-(E)-isomer 29 as product. The assignments of absolute configurations for the products from dialdehydes 14 have been made on the basis of similar investigations. The (E)-product 23a was converted to both diastereomers of the Mosher ester derivative 30, and NMR analysis24 of these compounds gave the assigned absolute configuration. The pivaloylprotected (E)-product 23b was correlated with 23a by conversion of both compounds to derivative 31; based on NMR analysis, the same diastereomer of 31 was produced in both cases. The (Z)-products 25a and 25b were correlated with 23a and 23b, respectively, by conversion to the mono-(Z)/mono-(E)-bisaddition products: 23a and 25a both formed 32a, and 23b and 25b both gave 32b. Discussion Our results show that the structure of the group R1 in the chiral phosphonates 3 controls not only the (E)/(Z)selectivity of the asymmetric HWE reactions but also which enantiotopic carbonyl group in the dialdehyde substrate reacts faster,25 both of which factors contribute to giving these reactions increased synthetic versatility. The bis(trifluoroethyl) reagent 3d was found to give (Z)-products with excellent diastereoselectivities in good yields from all three substrates studied. The results obtained with 14a show that the specific choice of reaction conditions is important: although the use of KHMDS/18-crown-6 as base often is the most efficient (24) (a) Dale, J. A.; Mosher, H. S. J. Am. Chem. Soc. 1973, 95, 512519. (b) Ohtani, I.; Kusumi, T.; Kashman, Y.; Kakisawa, H. J. Am. Chem. Soc. 1991, 113, 4092-4096, and references therein. For further details regarding the preparation and NMR analyses of compounds 27 and 30, see Supporting Information. (25) We have found the same general trend to be valid also in kinetic/dynamic resolutions of racemic monoaldehydes by reaction with phosphonates 3a-d; see refs 4o, 4s, and 4t.

Asymmetric Horner-Wadsworth-Emmons Reactions

J. Org. Chem., Vol. 63, No. 23, 1998 8289

Reactions of Phosphonates 3a-d with Dialdehydes 14a and 14ba

Table 3. entry

phosphonate

substrate

temp. (°C)

product

yieldb (%)

diast. ratioc

yield of bisaddition (%)

1 2 3 4 5 6 7 8 9 10

3a 3b 3c 3d 3dg 3di 3a 3b 3c 3d

14a 14a 14a 14a 14a 14a 14b 14b 14b 14b

-78 -78 -78 -78 -78 -78 -90 -90 -90 -90

23a 23a 23a 25a 25a 25a 23b 23b 23b 25b

65d 60d 46d 61f 76 60 10 62 65 62

g98:2 (87:13) g98:2 (88:12) g98:2 (91:9) 86:14 98:2 96:4 82:18 95:5 94:6 g98:2

e e e 24f h h 65j 35j 26j 29j

a General reaction conditions: 1.1-1.3 equiv of dialdehyde, 1.05-1.2 equiv of phosphonate, 1.0 equiv of KHMDS, 5 equiv of 18-crown6, ca. 0.02 M in THF, 6-17 h. b Isolated yield of product judged as g95% pure by NMR and TLC, unless otherwise noted. c Ratio in isolated product. If different, the ratio in the crude product is given in parentheses. Entries 1-3: ratio 23a:24a; entries 4-6: ratio 25a:26a; entries 7-9: ratio 23b:24b; entry 10: ratio 25b:26b. Geometric ratios were g98:2, unless stated otherwise. dThe isolated product also contained small amounts of unreacted dialdehyde: entry 1, 6%; entry 2, 6%; entry 3, 3%. e After chromatography, (E,E)-bis-addition product and the minor (E)-monoaddition diastereomer 24a were obtained as a mixture; the yield of bis-addition product was not determined. f Some (Z,E)-bis-addition product 32a (7% yield; see Chart 2) was obtained in the same fraction as the monoaddition product. The (Z,Z)bis-addition product was isolated separately (17% yield). g Only KHMDS (no 18-crown-6) used as base. h Not determined. i NaHMDS used as base. j Mixture of bis-addition products.

Chart 1

Scheme 7

Chart 2

with other substrates, the particular reaction between 3d and 14a gave best results if the crown ether was omitted. The (E)-selective reagents 3a-c, on the other hand, generally performed best in combination with KHMDS/18-crown-6. Depending on the substrate, either 3a or 3b proved to be the most efficient reagent, the diisopropyl phosphonate 3c seemingly being too sterically hindered to be generally useful. By choosing the appropriate reagent, (E)-products were obtained with di-

astereoseletivities g 95:5 and in synthetically useful yields. An additional factor to take into account is the choice of protecting group in R-oxygenated substrates: although substrates 14a and 14b gave the same overall trends in their reactions, the structure of the protecting group clearly influenced both the yields and selectivities of the reactions.26 As demonstrated by the results obtained in reactions with 6 (Table 1), the diastereoselectivity observed for the monoaddition product depends on the reaction stoichiometry. This outcome is to be expected, since increased conversion to the bisaddition product (e.g., 33, Scheme 7) should increase the ratio between the diastereomeric monoaddition products (the ratio 19:20 in the example shown), due to the minor monoaddition diastereomer reacting faster in the second step.27,28 (26) A topic worthy of future investigation is whether the use of another protecting group would enable a change in the mechanism by which the R-stereocenter in the substrate influences the reaction stereochemistry (see mechanistic discussion below, and also ref 4cc); such a change could, in turn, enable complementary preparation of other product diastereomers.

8290 J. Org. Chem., Vol. 63, No. 23, 1998

Mechanism. Recent results from computational studies, using both high level ab initio calculations29 and molecular mechanics methods,30 fully support our previously reported working model4a,cc for rationalizing the stereochemical outcome of these asymmetric HWE reactions. This analysis is based on the postulate that phosphonates 3 form (E)-enolates31 34 under the reaction conditions used, an assumption which is reasonable based on previous studies.32,33 Since two new stereocenters are formed at the bond-forming sites (C2 and C3) in the intermediates, and the substrate contains two enantiotopic carbonyl groups, eight different diastereomeric forms of the intermediate oxyanion 35 are theoretically possible. Our computational studies indicate that for reactions involving phosphonates 3a-c, which normally are (E)-selective, the transition states for oxyanion formation (TS1) and for ring closure to the oxaphosphetane 36 (TS2) are close in energy, making it necessary to include appropriate models of both transition states

when analyzing reaction stereoselectivities. In reactions with trifluoroethyl phosphonate 3d, on the other hand, the initial addition step will be irreversible, and it will be sufficient to model diastereomeric forms of TS1. The particular diastereomeric intermediates which according to our modeling studies are the precursors of the main product isomers are illustrated in Schemes 8 and 9. The formation of these intermediates is rationalized as follows. The chiral auxiliary efficiently blocks the Re-face of the phosphonate enolate 34. Diastereomers of TS1 leading to oxyanions with (2R)-configuration are therefore prohibitively high in energy, leading to a very high preference for oxyanions with (2S)-configuration. The configuration at the former aldehyde carbonyl carbon, C3, is mainly controlled by the stereochemistry at C4, the former aldehyde R-carbon. The effect can be interpreted as a formal Felkin-Anh-Eisenstein34,35 (FAE) or anti-FAE effect in TS1, depending on the substituents (27) Schreiber, S. S.; Schreiber, T. S.; Smith, D. B. J. Am. Chem. Soc. 1987, 109, 1525-1529, and references therein. (28) This expectation rests on the assumption that the unreacted aldehyde carbonyl groups in the monoaddition diastereomers show relative reactivities similar to the corresponding enantiotopic carbonyl groups in the dialdehyde substrate. (29) Brandt, P.; Norrby, P.-O.; Martin, I.; Rein, T. J. Org. Chem. 1998, 63, 1280-1289. (30) (a) Norrby, P.-O. In Transition State Modelling for Catalysis; Truhlar, D., Morokuma, K., Eds.; ACS Symposium Series, in press. (b) Norrby, P.-O.; Brandt, P.; Rein, T., manuscript in preparation. At present, modeling tools are available for reactions involving phosphonates containing simple alkyl groups (e.g., 3a-c). Work toward the design of parameter sets which will allow modeling of trifluoroethyl reagents is in progress. (31) Note that the designation of the geometry of enolate 34 as (E) or (Z) depends on whether a counterion is considered as being bonded to the anionic oxygen or not, and if so, on the CIP priority of that counterion relative to carbon. (32) Gais and co-workers have reported4y that the lithium enolate formed from 3a has (E)-geometry. We are presently studying the stuctures of enolates derived from reagents 3b-d, as well as the dependence of the enolate ratios on the reaction conditions; these studies will be reported upon separately. (33) (a) Bottin-Strzalko, T.; Corset, J.; Froment, F.; Pouet, M.-J.; Seyden-Penne, J.; Simmonin, M.-P. J. Org. Chem. 1980, 45, 12701276. (b) Seyden-Penne, J. Bull. Soc. Chim. France 1988 (II), 238242, and references therein.

Tullis et al. Scheme 8

Scheme 9

at C4, but is in fact even more pronounced in the tighter TS2. The overall effect is that oxyanion diastereomers with an unfavorable relative configuration at C3/C4 will revert to starting material instead of proceeding to product. This analysis also explains the, at first perhaps puzzling, observation that the major (E)- and (Z)-products have opposite absolute configuration at the allylic stereocenter. Since the favored absolute configuration at C2 always is the same, but the configuration at C3 is controlled by the one at C4, it automatically follows that reactions at opposite enantiotopic aldehyde groups must lead to intermediates with opposite relative configuration at C2/C3 and therefore to products with opposite alkene geometry.36 (34) (a) Che´rest, M.; Felkin, H. Tetrahedron Lett. 1968, 2205-2208. (b) Anh, N. T.; Eisenstein, O.; Lefour, J.-M.; Traˆn Huu Daˆu, M. E. J. Am. Chem. Soc. 1973, 95, 6146-6147. (c) Anh, N. T.; Eisenstein, O. Nouv. J. Chem. 1977, 1, 61-70. (d) Anh, N. T. Top. Curr. Chem. 1980, 88, 145-162. (e) Lodge, E. P.; Heathcock, C. H. J. Am. Chem. Soc. 1987, 109, 2819-2820. (f) Lodge, E. P.; Heathcock, C. H. J. Am. Chem. Soc. 1987, 109, 3353-3361. For excellent discussions of different models for diastereoselection in nucleophilic additions to carbonyl compounds, see: (g) Roush, W. R. J. Org. Chem. 1991, 56, 4151-4157. (h) Evans, D. A.; Dart, M. J.; Duffy, J. L.; Yang, M. G. J. Am. Chem. Soc. 1996, 118, 4322-4343. (i) Gung, B. W. Tetrahedron 1996, 52, 5263-5301. (35) Evans has introduced34h a stereochemical model which rationalizes the merged influence of R- and β-substituents in aldol-type additions to substituted aldehydes, including R-methyl-β-alkoxy-aldehydes. However, from substrates containing R- and β-substituents in an anti relationship (as in dialdehyde 6), the Evans model also predicts formation of FAE-type products.

Asymmetric Horner-Wadsworth-Emmons Reactions

It should also be noted that the favored relative configuration at C3/C4 cannot easily be predicted from empirical rules. However, our recently developed molecular mechanics method30 has been able to rationalize the product pattern. Conclusions In this paper, we have demonstrated that when mesodialdehydes 6 and 14 are used as substrates in asymmetric HWE reactions, either (E)- or (Z)-monoaddition products can be obtained with high geometric selectivity by slight structural variation in the phosphoryl group of the chiral reagent. By an appropriate choice of chiral reagent and reaction conditions, both (E)- and (Z)products are obtained with good to excellent diastereoselectivities and in synthetically useful yields; these results show that both R-methyl and R-oxygen substitution in the substrate enable high selectivities to be obtained. The reactions with dialdehydes 14 also show that the choice of oxygen protecting groups in the substrate can be used for fine-tuning the outcome. Furthermore, (E)- and (Z)-products are formed with opposite enantiotopic group preference. Thus, when applying these reactions in synthesis, both enantiomeric series of a projected synthetic intermediate can be accessed using the same enantiomer of the chiral auxiliary, as long as the alkene geometry of the product is of no consequence for the particular application at hand. The stereochemical outcome of these reactions is in all cases consistent with a mechanistic model in which the product stereochemistry is determined by the combined influence of the chiral auxiliary, the alkyl group in the phosphonate unit, and the R-stereocenters in the dialdehyde substrate. In particular, this model explains why reactions at opposite enantiotopic carbonyl groups in the substrate give products having opposite alkene geometry. Our continuing studies in this area are aimed at further improvement and generalization of this methodology through (1) the design of even more efficient chiral reagents, (2) investigations of mechanistic details, and (3) studies of reactions with new substrates. The products obtained from the asymmetric HWE reactions have projected utility in synthetic approaches to various natural products, and several such synthetic applications are under active investigation.

Experimental Section General. All reactions were performed in oven-dried or flame-dried glassware. Commercial reagents were used as received, unless otherwise indicated. Solvents were generally distilled before use. Potassium hexamethyldisilazide [KN(SiMe3)2, KHMDS] was purchased as a stock solution (0.5 M in toluene) and titrated according to the method of Ireland and Meissner.37 18-Crown-6 was recrystallized from anhydrous acetonitrile and dried under vacuum. Toluene, CH2Cl2, hexane, Et3N, and pyridine were distilled from CaH2. THF was distilled from sodium/benzophenone ketyl. Pd(OAc)2 was recrystallized from acetic acid, and benzoquinone was recrystallized from ethanol. Dialdehydes 6 and 14b were freshly (36) We have observed the same general trend in all asymmetric HWE reactions studied by us to date. A similar argumentation is used by Fuji and co-workers when analyzing the results of asymmetric HWE reactions between a chiral phosphonate and a group of meso-diketones; see ref 4bb. (37) Ireland, R. E.; Meissner, R. S. J. Org. Chem. 1991, 56, 45664568.

J. Org. Chem., Vol. 63, No. 23, 1998 8291 prepared just before use. Cooling below -78 °C was effected by use of either EtOH/liquid N2 or an immersion cooler. TLC analyses were performed on Merck aluminum-backed F254 silica gel plates, using UV light and phosphomolybdic acid for visualization. Drying of organic phases obtained from extractive workup was generally done with MgSO4. Flash chromatography was performed as described by Still and co-workers38 and medium-pressure liquid chromatography (MPLC) as described by Baeckstro¨m and co-workers,19 in both cases using either Merck silica gel 60 (230-400 mesh) or Amicon Matrex 60 Å silica gel (35-70 µm). NMR spectra were recorded in CDCl3 unless otherwise indicated, using CHCl3 (δ 7.24 ppm) and CDCl3 (δ 77.0 ppm) as internal references for 1H and 13C, respectively. (1R,2S,5R)-8-Phenylmenthol was prepared according to a literature procedure.39 Tropone40 and 1,3-cycloheptadien-6-ol41 are intermediates in the synthetic route to diene 9 and were prepared according to literature procedures. Stereochemical descriptors for the meso compounds reported here have been assigned in accordance with a recent treatise.42 (Diethoxyphosphoryl)acetic Acid (1R,2S,5R)-5-Methyl2-(1-methyl-1-phenylethyl)cyclohexyl Ester (3b). Prepared from triethyl phosphonoacetate (1b) and (1R,2S,5R)-8phenylmenthol (2) in 81% yield, in analogy with the published procedure8 for preparation of 3a. 1H NMR (400 MHz, selected data) δ 7.24-7.16 (m, 4 H), 7.08-7.03 (m, 1 H), 4.77 (ddd [app td], J ) 10.8, 4.4 Hz, 1 H), 4.08-3.91 (m, 4 H), 2.31 (dd, J ) 21.3, 14.4 Hz, 1 H), 2.02 (dd, J ) 21.3, 14.4 Hz, 1 H), 0.81 (d, J ) 6.5 Hz, 3 H); 13C NMR (100 MHz) δ 165.1 (d, J ) 5.8 Hz), 151.8, 127.9 (2 C), 125.4 (2 C), 125.1, 75.1, 62.4 (d, J ) 5.7 Hz, 2 C), 50.3, 41.3, 39.5, 34.5, 33.9 (d, J ) 133 Hz), 31.3, 29.2, 26.3, 23.2, 21.8, 16.3 (d, J ) 6.1 Hz, 2 C). Anal. Calcd for C22H35O5P: C, 64.37; H, 8.59. Found: C, 63.97; H, 8.51. [Bis(2,2,2-trifluoroethoxy)phosphoryl]acetic Acid (1R,2S,5R)-5-Methyl-2-(1-methyl-1-phenylethyl)cyclohexyl ester (3d). Prepared from methyl bis(trifluoroethyl) phosphonoacetate (1d) and (1R,2S,5R)-8-phenylmenthol in 89% yield, according to the same procedure as was used for 3b. 1H NMR (400 MHz, selected data) δ 7.30-7.23 (m, 4 H), 7.17-7.07 (m, 1 H), 4.82 (ddd [app td], J ) 10.7, 4.4 Hz, 1 H), 4.48-4.20 (m, 4 H), 2.27 (dd, J ) 20.7, 15.9 Hz, 1 H), 2.25 (dd, J ) 20.4, 15.9 Hz, 1 H), 1.27 (s, 3 H), 1.17 (s, 3 H), 0.86 (d, J ) 6.4 Hz, 3 H); 13C NMR (100 MHz) δ 164.0 (d, J ) 5.8 Hz), 151.9, 128.1 (2 C), 125.3 (2 C), 125.1, 122.4 (app quartet of m, one J ) ca 277 Hz, 2 C), 75.9, 62.4 (app quartet of m, one J ) ca. 40 Hz, 2 C), 50.1, 41.2, 39.3, 34.4, 33.4 (d, J ) 144 Hz), 31.2, 30.0, 26.1, 22.2, 21.7. Anal. Calcd for C22H29O5FP: C, 50.97; H, 5.64. Found: C, 50.81; H, 5.66. meso-(2R,3r,4S)-3-((tert-(Butyldimethylsilyl)oxy)-2,4dimethylpentanedial (6). To a solution of oxalyl chloride (136 µL, 1.55 mmol) in CH2Cl2 (8 mL) at -78 °C under argon was added dropwise a solution of DMSO (147 µL, 2.07 mmol) in CH2Cl2 (1 mL). After 30 min, a solution of 514 (136 mg, 0.518 mmol) in CH2Cl2 (1 mL) was added dropwise followed, after 30 min, by triethylamine (722 µL, 5.18 mmol). The reaction mixture was stirred at -78 °C for 1 h and then warmed slowly to 0 °C during 1 h. After 30 min of stirring at 0 °C, the solution was diluted with dry toluene (25 mL), filtered through a dried glass frit, and concentrated. The residue was dissolved in dry hexane (25 mL), filtered again, and concentrated yielding 131 mg (quantitative crude yield) of 6 as a pale yellow oil which due to its limited stability was used without further purification in the HWE reactions: 1H NMR (250 MHz) δ 9.73 (d, J ) 2.1 Hz, 2H), 4.28 (t, J ) 5.1 Hz, 1H), 2.59 (qdd, J ) 7.1, 5.1, 2.0 Hz, 2H), 1.08 (d, J ) 7.1 Hz, 6H), 0.84 (s, 9H), 0.06 (s, 6H); 13C NMR (62.5 MHz) δ 203.4, 74.2, 50.8, 25.7, 18.0, 10.5, -4.5. (38) Still, W. C.; Kahn, M.; Mitra, A. J. Org. Chem. 1978, 43, 29232925. (39) Ort, O. Org. Synth. 1987, 65, 203-214. (40) Dahnke, K. R.; Paquette, L. A. Org. Synth. 1992, 71, 181-188. (41) Schuster, D. I.; Palmer, J. M.; Dickerman, S. C. J. Org. Chem. 1966, 31, 4281-4282. (42) Helmchen, G. In Houben-Weyl, Stereoselective Synthesis; Helmchen, G., Hoffman, R. W., Mulzer, J., Schaumann, E., Eds.; Thieme: Stuttgart, 1996; Vol. 1, pp 1-74.

8292 J. Org. Chem., Vol. 63, No. 23, 1998 meso-(3R,5s,7S)-5-(Benzyloxy)-3,7-diacetoxycyclohept1-ene (10). Diene 917 (1.00 g, 5.0 mmol) was added neat to a solution of palladium(II) acetate (56 mg, 0.25 mmol), lithium acetate dihydrate (2.55 g, 25 mmol), and benzoquinone (108 mg, 1.0 mmol) in glacial acetic acid (10 mL). After addition of MnO2 (869 mg, 10 mmol), the resulting slurry was stirred at room temperature. The reaction was followed by TLC until all starting material had been consumed (2-3 days) and then worked up by addition of water and hexane/EtOAc 1/1. The resulting emulsion was filtered through Celite, the phases were separated, and the aqueous phase was extracted with three portions of hexane/EtOAc 1/1. The combined organic phases were washed with water, 2 M NaOH, and water again and then dried and concentrated to give 1.42 g of a yellowishbrown oil. Purification was effected by MPLC (0-100% EtOAc in hexanes), yielding 811 mg (51%) of 10 as a colorless oil: 1H NMR (250 MHz) δ 7.42-7.23 (5H), 5.76 (dd, J ) 10.8, 2.3 Hz, 2H), 5.7 (s, 2 H), 4.65 (s, 2H), 3.92 (tt, J ) 5.6, 2.7 Hz, 1H), 2.22 (ddd, J ) 13.4, 5.6, 2.3 Hz, 2H), 1.86 (ddd, J ) 1.4, 10.8, 2.7 Hz, 2H), 2.05 (s, 6H); 13C NMR (62.5 MHz) δ 169.9, 138.3, 132.5, 128.2, 127.5, 127.3, 71.7, 70.0, 68.8, 36.6, 21.1. Anal. Calcd for C18H22O5: C, 67.91; H, 6.97. Found: C, 67.83; H, 6.85. meso-(1R,4S,6s)-6-(Benzyloxy)cyclohept-2-ene-1,4-diol (11). To a solution of diacetate 10 (111 mg, 0.35 mmol) in MeOH (1.5 mL) at 0 °C was added dropwise an aqueous solution of KOH (98 mg, 1.75 mmol, in 1.5 mL of water). The resulting pale yellow solution was stirred at 0 °C until no starting material was detected by TLC (ca. 30 min). The reaction mixture was neutralized with 2 M H2SO4 and diluted with water. Extractive workup (CH2Cl2), drying, and concentration gave 82 mg of a white solid. Purification by flash chromatography (hexanes/EtOAc 1/1) gave 76 mg (93%) of white crystalline 11: 1H NMR (400 MHz, DMSO) δ 7.37-7.23 (m, 5H), 5.55 (s, 2H), 4.79 (d, J ) 4.7, 2H), 4.53 (s, 2H), 4.504.41 (m, 2H), 3.83-3.75 (m, 1H), 2.02 (ddd, J ) 13.3, 5.2, 2.4 Hz, 2H), 1.55 (ddd, J ) 13.4, 11.1, 2.5 Hz, 2H); 13C NMR (100 MHz, DMSO) δ 139.0, 136.1, 128.3, 127.4, 127.3, 73.2, 69.1, 64.1, 40.0. Anal. Calcd for C14H18O3: C, 71.77; H, 7.74. Found: C, 71.88; H, 7.61. meso-(3R,5s,7S)-5-(Benzyloxy)-3,7-bis-((tert-butyldiphenylsilyl)oxy)cyclohept-1-ene (12a). To a solution of diol 11 (0.816 g, 3.35 mmol) and imidazole (1.14 g, 16.75 mmol) in dry DMF (30 mL) under argon was added tert-butyldiphenylsilyl chloride (2.1 mL, 8.05 mmol). The reaction mixture was stirred for 24 h at room temperature and then diluted with brine and extracted with ethyl acetate. Drying (Na2SO4) and concentration of the combined organic extracts followed by flash chromatography (hexanes/EtOAc 99/1) afforded 2.432 g (99%) of 12a as a colorless oil: 1H NMR (400 MHz) δ 7.727.64 (m, 8H), 7.44-7.23 (m, 15H), 7.00-6.93 (m, 2H), 5.78 (s, 2H), 4.75 (d, J ) 9.9 Hz, 2H), 3.89 (s, 2H), 3.67-3.60 (m, 1H), 2.02 (ddd [app br dd], J ) 12, 4 Hz, 2H), 1.78 (ddd [app br t], J ) 12 Hz, 2H), 1.09 (s, 18H); 13C NMR (100 MHz) δ 138.9, 135.8, 134.3, 134.0, 129.5, 128.0, 127.9, 127.5, 127.0, 72.6, 67.5, 69.0, 40.1, 27.0, 26.49, 19.2. Anal. Calcd for C46H54O3Si2: C, 77.70; H, 7.65. Found: C, 77.48; H, 7.62. meso-(1R,2S,3R,5s,7S)-5-(Benzyloxy)-3,7-bis((tert-butyldiphenylsilyl)oxy)cycloheptane-1,2-diol and meso(1R,2S,3S,5s,7R)-5-(Benzyloxy)-3,7-bis-(tert-butyldiphenylsilyl)oxy)cycloheptane-1,2-diol (13a; mixture of the two meso diols). To a solution of silyl ether 12a (321 mg, 0.45 mmol) and N-methylmorpholine N-oxide (53 mg, 0.45 mmol) in 5 mL of THF were added tert-butyl alcohol (2.5 mL) and H2O (1.2 mL), followed by 115 µL of a 2.5 weight-% solution of OsO4 in tert-butyl alcohol (0.009 mmol). The pale yellow solution was stirred for 30 h at room temperature and then quenched with 5 mL of 10% Na2S2O4. Dilution with brine followed by extraction (EtOAc), drying, and concentration followed by flash chromatography (hexanes/EtOAc 88/12) afforded 325 mg of a mixture of diastereomeric diols 13a (97%) as a colorless oil. The diastereomeric mixture was normally used as such for preparation of the dialdehyde 14a; isomerically pure samples of the separate diastereomeric diols could be obtained by chromatography, however. Faster eluting

Tullis et al. isomer (Rf ) 0.42, hexanes/EtOAc 8/2): 1H NMR (250 MHz) δ 7.75-7.60 (m, 8H), 7.46-7.20 (m, 15H), 6.97-6.90 (m, 2H), 4.06 (br d, J ) 10.2 Hz, 2H), 3.91 (s, 2H), 3.65 (br s, 2H), 3.533.43 (m, 1H), 3.10 (br d, J ) 4.9 Hz, 2H), 2.15 (ddd, J ) 14.4, 10.1, 4.1 Hz, 2H), 1.68 (ddd, J ) 14.4, 5.8, 2.8 Hz, 2H), 1.06 (s, 18H); 13C NMR (62.5 MHz) δ 138.6, 135.9, 133.6, 133.3, 129.8, 128.1, 127.7, 127.2, 74.7, 71.4, 70.6, 69.6, 34.8, 27.0, 19.2. Anal. Calcd for C46H56O5Si2: C, 74.15; H, 7.58. Found: C, 74.37; H, 7.78. Slower eluting isomer (Rf ) 0.28, hexanes/ EtOAc 8/2): 1H NMR (250 MHz) δ 7.75-7.65 (m, 8H), 7.507.20 (m, 15H), 7.08-7.00 (m, 2H), 4.21 (br td, J ) 7, 3 Hz, 2H), 4.01 (br dd, J ) 5.9, 2.2 Hz, 2H), 3.96 (s, 2H), 3.70-3.58 (m, 1H), 2.94 (d, J ) 2.4 Hz, 2H), 1.97 (ddd, J ) 14.7, 6.4, 3.2 Hz, 2H), 1.88 (ddd, J ) 14.7, 7.6, 4.8 Hz, 2H), 1.11 (s, 18H); 13C NMR (62.5 MHz) δ 138.3, 135.9, 133.9, 133.3, 129.9, 128.2, 127.9, 127.8, 127.4, 74.7, 71.0, 70.2, 69.7, 36.2, 27.1, 19.4. Anal. Calcd for C46H56O5Si2: C, 74.15; H, 7.58. Found: C, 74.50; H, 7.82. meso-(2R,4s,6S)-4-(Benzyloxy)-2,6-bis(tert-butyldiphenylsilyl)oxy)heptanedial (14a). To a solution of diols 13a (412 mg, 0.552 mmol) in THF (10 mL) was added a solution of H5IO6 (125 mg, 0.552 mmol) in THF (10 mL). The reaction mixture became cloudy after a few minutes. After 2.5 h at room temperature, 20 mL of pH 7 phosphate buffer was added, and the solution was extracted with ethyl acetate. Drying and concentration followed by flash chromatography (hexanes/EtOAc 9/1) yielded 365 mg (89%) of 14a as a pale beige oil: 1H NMR (400 MHz) δ 9.50 (d, J ) 1.7 Hz, 2H), 7.687.58 (m, 8H), 7.46-7.22 (m, 15H), 7.11-7.05 (m, 2H), 4.15 (ddd [app td], J ) 6.0, 1.5 Hz, 2H), 4.06 (s, 2H), 3.75-3.67 (m, 1H), 2.02 (ddd [app td], J ) 14.6, 6.2 Hz, 2H), 1.78 (ddd, J ) 14.6, 6.0, 4.9 Hz, 2H), 1.11 (s, 18H); 13C NMR (100 MHz, some signals in the aromatic region overlap) δ 202.2, 137.8, 135.8, 135.8, 132.9, 132.8, 130.1, 128.2, 127.8, 127.7, 127.4, 76.0, 72.1, 70.1, 38.3, 26.9, 19.3. Anal. Calcd for C46H54O5Si2: C, 74.35; H, 7.32. Found: C, 74.12; H, 7.16. meso-(3R,5s,7S)-5-(Benzyloxy)-3,7-bis((2,2-dimethylpropionyl)oxy)cyclohept-1-ene (12b). To a solution of diol 11 (442 mg, 1.39 mmol) and DMAP (170 mg, 1.39 mmol) in dry pyridine (25 mL) under argon was added pivaloyl chloride (520 µL, 4.22 mmol). The reaction mixture was stirred for 16 h at reflux. After cooling and concentration, the residue was purified by flash chromatography (hexanes/EtOAc 95/5) to afford 555 mg (99%) of 12b as white solid: 1H NMR (400 MHz) δ 7.39-7.21 (m, 8H), 5.71 (br dd, J ) 11, 3 Hz, 2H), 5.66 (s, 2H), 4.63 (s, 2H), 3.9 (tt [app septet], J ) 5.8, 2.9 Hz, 1H), 2.18 (ddd, J ) 13.4, 5.8, 2.7 Hz, 2H), 1.89 (ddd, J ) 13.4, 10.7, 2.7 Hz, 2H), 1.18 (s, 18H); 13C NMR (100 MHz) δ 177.2, 138.3, 132.5, 128.2, 127.5, 127.3, 71.6, 70.0, 68.5, 38.4, 36.7, 27.0. Anal. Calcd for C24H34O5: C, 71.61; H, 8.51. Found: C, 71.61; H, 8.49. meso-(1R,2S,3R,5s,7S)-5-(Benzyloxy)-3,7-bis((2,2-dimethylpropionyl)oxy)cycloheptane-1,2-diol and meso(1R,2S,3S,5s,7R)-5-(Benzyloxy)-3,7-bis(2,2-dimethylpropionyl)oxy)cycloheptane-1,2-diol (13b; mixture of the two meso diols). To a solution of bis-pivaloyl ester 12b (557 mg, 1.38 mmol) and N-methylmorpholine N-oxide (162 mg, 1.38 mmol) in 16 mL of THF were added tert-butyl alcohol (8 mL) and H2O (4 mL), followed by 325 µL of a 2.5 wt % solution of OsO4 in tert-butyl alcohol (0.026 mmol). The reaction mixture was stirred for 4 h at room temperature and then quenched with 7 mL of 10% Na2S2O4. Dilution with brine followed by extraction (EtOAc), drying, and concentration afforded an essentially quantitative yield of a crude mixture of diols 13b as a colorless oil. This diastereomeric mixture could be used in the preparation of dialdehyde 14b without further purification. Alternatively, isomerically pure samples of the separate diol diastereomers could be obtained by flash chromatography38 (hexanes/EtOAc 7/3). Faster eluting isomer (Rf ) 0.36, hexanes/EtOAc 7/3): 1H NMR (400 MHz) δ 7.337.22 (m, 5H), 5.19 (dm, J ) 10.1 Hz, 2H), 4.49 (s, 2H), 4.03 (br s, 2H), 3.82 (tt, J ) 6.5, 4.4 Hz, 1H), 2.81 (br s, 2H), 2.37 (ddd, J ) 14.3, 10.1, 4.6 Hz, 2H), 1.94 (ddd, J ) 14.3, 6.4, 3.1 Hz, 2H), 1.18 (s, 18H); 13C NMR (100 MHz) δ 177.9, 138.2, 128.3, 127.5, 73.2, 71.0, 70.5, 70.3, 38.8, 31.9, 27.0. Slower eluting

Asymmetric Horner-Wadsworth-Emmons Reactions isomer (Rf ) 0.22, hexanes/EtOAc 7/3): 1H NMR (400 MHz) δ 7.31-7.24 (m, 5H), 5.15 (ddd [app td], J ) 6.4, 4.4 Hz, 2H), 4.54 (s, 2H), 3.91 (d, J ) 6.4 Hz, 2H), 3.71 (tt, J ) 7.8, 2.9 Hz, 1H), 3.44 (br s, 2H), 2.30 (ddd, J ) 15.0, 7.6, 4.3 Hz, 2H), 1.95 (ddd, J ) 15.0, 6.1, 3.05 Hz, 2H), 1.17 (s, 18H); 13C NMR (100 MHz) δ 178.1, 137.6, 128.4, 127.8, 127.7, 73.5, 71.8, 70.4, 70.2, 38.7, 35.1, 27.0. Anal. Calcd for C24H36O7: C, 66.03; H, 8.31. Found: C, 66.02; H, 8.37. meso-(2R,4s,6S)-4-(Benzyloxy)-2,6-bis(2,2-dimethylpropionyl)oxy)heptanedial (14b). To a solution of diols 13b (147 mg, 0.339 mmol) in THF (4 mL) was added a solution of H5IO6 (77.2 mg, 0.339 mmol) in THF (4 mL) at 0 °C. The ice bath was removed after 15 min; the reaction mixture then became cloudy in 10 min. After 1.5 h at room temperature, 3 mL of pH 7 phosphate buffer and 10 mL of brine were added. The solution was extracted with EtOAc, dried, and concentrated, and the residue was dissolved in CHCl3 (2 mL) and filtered through a plug of cotton, affording 147 mg (99%) of a colorless oil after concentration. Dialdehyde 14b was used in HWE reactions without further purification: 1H NMR (400 MHz) δ 9.48 (s, 2H), 7.37-7.27 (m, 5H), 5.13 (dd, J ) 9.8, 3.7 Hz, 2H), 4.38 (s, 2H), 3.69 (tt [app sept], J ) 7.9, 4.0, 1H), 2.18 (ddd, J ) 15.0, 7.9, 3.4 Hz, 2H), 1.95 (ddd, J ) 15.0, 9.5, 4.0 Hz, 2H), 1.22 (s, 18H); 13C NMR (100 MHz) δ 197.5, 177.8, 137.1, 128.6, 128.1, 75.3, 72.2, 71.5, 38.7, 34.0, 27.0. General Procedure for the Asymmetric HWE Reactions. To a solution of the chiral phosphonate (3a-d; 1.051.2 equiv) and 18-crown-6 (if applicable; 5 equiv) in THF (ca 0.02 M with respect to the phosphonate) at -78 °C under argon was added 1.0 equiv of KHMDS (0.5 M in toluene). After 30 min the resulting grayish slurry was transferred via cannula to a precooled solution of the dialdehyde43 (6, 14a or 14b; Table 1: 1.2-2.1 equiv, Tables 2 and 3: 1.1-1.3 equiv). The reaction mixture was stirred for the indicated time at the appropriate reaction temperature (monitoring by TLC) and then quenched with acetic acid (1 M in MeOH or THF). After 5 min, pH 7 phosphate buffer was added, and the reaction was slowly warmed to room temperature. After dilution with water, extractive workup (ethyl acetate), drying, and concentration gave the crude condensation products. Purification by flash chromatography or MPLC19 using EtOAc in hexanes as eluent afforded the products as colorless oils.23 General Procedure for Reduction of the HWE Products. The crude condensation product was dissolved in MeOH or i-PrOH (THF was sometimes added to increase solubility) at 0 °C, and NaBH4 (5-10 equiv) was added. After stirring at 0 °C until the reaction was finished (monitoring by TLC), the reaction mixture was diluted with water and extracted with CH2CH2. Drying, concentration, and purification by flash chromatography (EtOAc in hexanes) gave the alcohols as colorless oils. (E)-(4S,5S,6R)-5-(tert-(Butyldimethylsilyl)oxy)-7-hydroxy-4,6-dimethylhept-2-enoic Acid (1R,2S,5R)-5-Methyl-2-(1-methyl-1-phenylethyl)cyclohexyl Ester (15). Prepared from 3a and 6 in 77% overall yield; (E):(Z) ) 98:2, diastereomeric ratio 15:16 ) 91:9.23 15: 1H NMR (250 MHz, selected data) δ 7.27-7.19 (m, 4H), 7.13-7.06 (m, 1H), 6.72 (dd, J ) 15.8, 7.7 Hz, 1H), 5.21 (dd, J ) 15.7, 1.3 Hz, 1H), 4.84 (ddd [app td], J ) 10.7, 4.3 Hz, 1H), 3.61-3.53 (br m, 2H), 3.57 (dd [app t], J ) 5.1 Hz, 1H), 2.56-2.43 (m, 1H), 2.31 (br s, 1H), 2.00 (ddd, J ) 12.1, 10.7, 3.3 Hz, 1H), 1.84-1.73 (m, 1H), 1.28 (s, 3H), 1.19 (s, 3H), 1.00 (d, J ) 6.9 Hz, 3H), 0.90 (d, J ) 6.8 Hz, 3H), 0.89 (s, 9H), 0.84 (d, J ) 6.5 Hz, 3H), 0.09 (s, 3H), 0.06 (s, 3H); 13C NMR (100 MHz) δ 165.7, 151.6, 150.4, 127.9, 125.4, 124.9, 121.5, 79.8, 74.3, 65.7, 50.6, 41.72, 39.8, 38.0, 34.6, 31.3, 27.2, 26.7, 26.0, 25.7, 21.8, 18.2, 15.6, 15.6, -4.09, -4.12. Anal. Calcd for C31H52O4Si: C, 72.04; H, 10.14. Found: C, 71.81; H, 10.13. 16: 1H NMR (250 MHz, selected data assigned from a mixture (91:9) of diastereomers 15 and 16) δ 6.70 (dd, J ) 15.8, 7.4 Hz, 1H), 5.18 (dd, J ) 15.8, 1.4 Hz, 1H). (43) The opposite mode of addition (i.e., adding a precooled solution of the aldehyde to the phosphonate enolate) gave identical results, if the transfer was performed rapidly.

J. Org. Chem., Vol. 63, No. 23, 1998 8293 (E)-(4R,6R,8S)-6-(Benzyloxy)-4,8-bis((tert-butyldiphenylsilyl)oxy)-9-hydroxynon-2-enoic Acid (1R,2S,5R)5-Methyl-2-(1-methyl-1-phenylethyl)cyclohexyl ester (17). Prepared from 3a and 14a in 60% overall yield; (E):(Z) ) 98: 2, diastereomeric ratio 17:18 ) 89:11.23 17: 1H NMR (400 MHz, selected data) δ 6.62 (dd, J ) 15.6, 5.9 Hz, 1H), 5.45 (dd, J ) 15.6, 1.1 Hz, 1H), 4.83 (ddd [app td], J ) 10.7, 4.3 Hz, 1H), 4.21 (app br q, J ) 5.8 Hz, 1H), 4.07 (d, J ) 11.1 Hz, 1H), 4.01 (d, J ) 11.1 Hz, 1H), 3.77-3.70 (m, 1H), 3.46 (ddd, J ) 11.6, 5.7, 3.5 Hz, 1H), 3.31 (ddd, J ) 11.6, 7.23, 4.3 Hz, 1H), 3.24 (dddd [app quintet], J ) ca 6 Hz, 1H), 1.27 (s, 3H), 1.24 (s, 3H), 1.05 (s, 9H), 1.02 (s, 9H), 0.87 (d, J ) 6.4 Hz, 3H); 13C NMR (100 MHz, some signals in the aromatic region overlap) δ 165.5, 150.8, 149.3, 138.2, 135.9, 135.7, 133.8, 133.6, 133.5, 133.2, 129.9, 129.8, 128.2, 127.9, 127.82, 127.76, 127.65, 127.57, 127.5, 127.4, 125.6, 125.2, 120.8, 74.5, 72.6, 71.6, 69.99, 69.95, 66.2, 50.7, 42.8, 41.8, 40.1, 38.7, 34.5, 31.3, 28.5, 27.04, 26.98, 25.0, 21.8, 19.3. Anal. Calcd for C64H80O6Si2: C, 76.76; H, 8.05. Found: C, 76.54; H, 7.96. 18: 1H NMR (400 MHz, selected data assigned from a mixture (89:11) of diastereomers 17 and 18) δ 6.41 (dd, J ) 15.6, 6.1 Hz, 1H), 5.34 (dd, J ) 15.6, 1.1 Hz, 1H). (E)-(4S,5S,6S)-5-(tert-(Butyldimethylsilyl)oxy)-4,6-dimethyl-7-oxo-hept-2-enoic Acid (1R,2S,5R)-5-Methyl-2-(1methyl-1-phenylethyl)cyclohexyl Ester (19). Prepared from 3a and 6 in 53% yield; (E):(Z) ) 98:2, diastereomeric ratio 19:20 ) 95:5.23,44 19: 1H NMR (250 MHz, selected data) δ 9.73 (d, J ) 2.5 Hz, 1H), 7.27-7.19 (m, 4H), 7.14-7.05 (m, 1H), 6.68 (dd, J ) 15.8, 7.7 Hz, 1H), 5.20 (dd, J ) 15.8, 1.3 Hz, 1H), 4. 82 (ddd [app td]), J ) 10.7, 4.3 Hz, 1H), 3.81 (dd [app t], J ) 4.6 Hz, 1H), 2.52-2.39 (m, 2H), 2.01 (ddd, J ) 12.1, 10.6, 3.4 Hz, 1H), 1.28 (s, 3H), 1.19 (s, 3H), 1.05 (d, J ) 7.1 Hz, 3H), 0.97 (d, J ) 6.9 Hz, 3H), 0.87 (s, 9H), 0.84 (d, J ) 6.5 Hz, 3H), 0.05 (s, 3H), 0.04 (s, 3H); 13C NMR (62.5 MHz) δ 204.3, 165.6, 151.7, 149.2, 127.9, 125.4, 124.9, 122.2, 77.2, 74.4, 50.5, 50.2, 41.7, 41.6, 39.7, 34.6, 31.3, 27.5, 26.6, 25.9, 25.4, 21.8, 18.1, 15.5, 11.7, -4.2, -4.4. Anal. Calcd for C31H50O4Si: C, 72.32; H, 9.79. Found: C, 72.06; H, 9.74. 20: 1H NMR (250 MHz, selected data assigned from a mixture (95: 5) of diastereomers 19 and 20) δ 6.66 (dd, J ) 16, 7 Hz, 1H), 5.17 (dd, J ) 16, 1.5 Hz, 1H). (Z)-(4R,5R,6R)-5-(tert-(Butyldimethylsilyl)oxy)-4,6dimethyl-7-oxo-hept-2-enoic Acid (1R,2S,5R)-5-Methyl2-(1-methyl-1-phenylethyl)cyclohexyl Ester (21). Prepared from 3d and 6 in 83% yield; (Z):(E) ) 98:2, diastereomeric purity ) 98:2. 21: 1H NMR (400 MHz, selected data) δ 9.74 (d, J ) 2.7 Hz, 1H), 7.25-7.20 (m, 4 H), 7.13-7.09 (m, 1H), 6.11 (dd, J ) 11.6, 10.0 Hz, 1H), 5.18 (dd, J ) 11.5, 0.6 Hz, 1H), 4.79 (ddd [app td], J ) 10.7, 4.3 Hz, 1H), 3.81 (dd, J ) 5.8, 3.2 Hz, 1H), 3.82 (m, 1H), 2.47 (qdd, J ) 7.0, 5.8, 2.7 Hz, 1H), 1.98 (ddd, J ) 12.3, 10.6, 3.2 Hz, 1H), 1.28 (s, 3H), 1.21 (s, 3H), 1.08 (d, J ) 7.1 Hz, 3H), 1.00 (d, J ) 6.9 Hz, 3H), 0.89 (s, 9 H), 0.84 (d, J ) 6.5 Hz, 3H), 0.09 (s, 3H), 0.05 (s, 3H); 13C NMR (100 MHz) δ 204.4, 165.1, 151.5, 149.9, 127.8, 125.4, 125.0, 120.4, 77.2, 73.9, 51.5, 50.5, 41.8, 39.7, 36.6, 34.6, 31.3, 26.9, 26.7, 25.8, 25.4, 21.6, 18.2, 17.6, 11.3, -4.1, -4.4. Anal. Calcd for C31H50O4Si: C, 72.32; H, 9.79. Found: C, 72.14; H, 9.63. (E)-(4R,6R,8S)-6-(Benzyloxy)-4,8-bis((tert-butyldiphenylsilyl)oxy)-9-oxonon-2-enoic Acid (1R,2S,5R)-5Methyl-2-(1-methyl-1-phenylethyl)cyclohexyl Ester (23a). Prepared from 3a and 14a in 65% yield;45 (E):(Z) ) 98:2, diastereomeric ratio 23a:24a ) 98:2 (in the crude product, this diastereomer ratio was 89:11, but the diastereomers could be separated by chromatography23). 23a: 1H NMR (400 MHz, selected data) δ 9.45 (d, J ) 1.7 Hz, 1H), 7.65-7.53 (m, 8 H), (44) The sample also contained 3% of an isomer tentatively assigned as being epimeric at the carbon R to the aldehyde carbonyl; see Supporting Information for details. (45) The sample obtained after flash chromatography also contained 6% of unreacted dialdehyde 14a; this could be removed, with good material recovery, by a second chromatography. Alternatively, it was easily removed by chromatography after NaBH4 reduction of the HWE product.

8294 J. Org. Chem., Vol. 63, No. 23, 1998 7.43-7.18 (m, 19H), 7.07-7.01 (m, 3H), 6.68 (dd, J ) 15.6, 6.1 Hz, 1H), 5.44 (dd, J ) 15.6, 1.2 Hz, 1H), 4.82 (ddd [app td], J ) 10.6, 4.3 Hz, 1H), 4.28 (app br q, J ) 6.2 Hz, 1H), 4.08 (ddd [app td], J ) 6.0, 1.8 Hz, 1H), 4.03 (d, J ) 11.1 Hz, 1H), 3.99 (d, J ) 11.0 Hz, 1H), 3.54 (dddd [app br quintet], J ) ca 5 Hz, 1H), 1.26 (s, 3H), 1.23 (s, 3H), 1.08 (s, 9H), 1.04 (s, 9H), 0.86 (d, J ) 6.5 Hz, 3H); 13C NMR (100 MHz, some signals overlap) δ 202.1, 165.5, 150.9, 149.1, 138.2, 135.97, 135.95, 135.92, 135.8, 133.6, 133.3, 133.0, 132.9, 130.2, 129.9, 128.3, 128.0, 127.9, 127.8, 127.7, 127.6, 127.4, 125.6, 125.5, 125.3, 121.1, 76.0, 74.6, 72.1, 70.3, 70.1, 50.7, 43.0, 41.8, 40.2, 38.2, 34.6, 31.4, 28.4, 27.1, 25.2, 21.9, 19.4. Anal. Calcd for C64H78O6Si2: C, 76.91; H, 7.87. Found: C, 76.63; H, 7.77. A sample enriched in diastereomer 24a was obtained from a different experiment. 24a: 1H NMR (400 MHz, selected data assigned from a mixture (87:13) of diastereomers 23a and 24a) δ 9.42 (d, J ) 1.8 Hz, 1H), 6.41 (dd, J ) 15.6, 6.1 Hz, 1H), 5.38 (dd, J ) 15.6, 1.4 Hz, 1H). (Z)-(4S,6S,8R)-6-(Benzyloxy)-4,8-bis((tert-butyldiphenylsilyl)oxy)-9-oxonon-2-enoic Acid (1R,2S,5R)-5Methyl-2-(1-methyl-1-phenylethyl)cyclohexyl Ester (25a). Prepared from 3d and 14a, using only KHMDS as base (i.e., no 18-crown-6), in 76% yield; (Z):(E) ) 98:2, diastereomeric ratio 25a:26a ) 98:2. 25a: 1H NMR (400 MHz, selected data) δ 9.41 (d, J ) 2.5 Hz, 1H), 7.66-7.51 (m, 8H), 7.44-7.01 (m, 22H), 5.95 (dd, J ) 11.6, 8.1 Hz, 1H), 5.44 (ddd [app br quartet], J ) ca 6 Hz, 1H), 4.75 (dd, J ) 11.6, 1.1 Hz, 1H), 4.56 (ddd [app td], J ) 10.6, 4.2 Hz, 1H), 4.26 (d, J ) 11.2 Hz, 1H), 4.23 (ddd [app td], J ) 6.4, 2.5 Hz, 1H), 4.11 (d, J ) 11.0 Hz, 1H), 3.75-3.66 (m, 1H), 1.12 (s, 3H), 1.10 (s, 12H), 1.00 (s, 9H), 0.89 (d, J ) 6.4 Hz, 3H); 13C NMR (100 MHz, some signals overlapping) δ 201.8, 164.3, 151.5, 151.1, 138.5, 135.8, 134.0, 133.7, 133.3, 133.2, 129.93, 129.87, 129.7, 129.6, 128.1, 127.8, 127.7, 127.6, 127.4, 127.2, 125.4, 124.9, 118.5, 76.5, 73.9, 72.1, 69.6, 67.8, 50.4, 42.1, 41.7, 39.8, 39.6, 34.5, 31.3, 27.3, 27.0, 26.6, 25.4, 21.9, 19.4, 19.2. Anal. Calcd for C64H78O6Si2: C, 76.91; H, 7.87. Found: C, 77.17; H, 7.58. A sample enriched in diastereomer 26a was obtained from a different experiment. 26a: 1H NMR (400 MHz, selected data assigned from a mixture (86:14) of diastereomers 25a and 26a) δ 9.38 (d, J ) 2.3 Hz, 1H), 6.00 (dd, J ) 11.6, 7.9 Hz, 1H), 4.93 (dd, J ) 11.7, 1.3 Hz, 1H). (E)-(4R,6R,8S)-6-(Benzyloxy)-4,8-bis((2,2-dimethylpropionyl)oxy)-9-oxonon-2-enoic Acid (1R,2S,5R)-5-Methyl2-(1-methyl-1-phenylethyl)cyclohexyl Ester (23b).46 Prepared from 3b and 14b in 62% yield; (E):(Z) ) 98:2, diastereomeric ratio 23b:24b ) 95:5.23 23b: 1H NMR (400 MHz, selected data) δ 9.47 (s, 1H), 7.38-7.17 (m, 9H), 7.117.02 (m, 1H), 6.49 (dd, J ) 15.8, 5.3 Hz, 1H), 5.50 (dddd, J ) 9.3, 5.0, 3.5, 1.3 Hz, 1H), 5.41 (dd, J ) 15.8, 1.5 Hz, 1H), 5.17 (dd, J ) 9.7, 3.5, 1H), 4.83 (ddd [app td], J ) 10.7, 4.3 Hz, 1H), 4.41 (d, J ) 10.6 Hz, 1H), 4.35 (d, J ) 10.5 Hz, 1H), 3.59 (46) Due to their limited stability, aldehydes 23b and 25b did not give satisfactory elemental analyses. The corresponding alcohols, obtained after NaBH4 reduction, were fully characterized, however; see Supporting Information for details.

Tullis et al. (dddd [app septet], J ) 4 Hz, 1H), 1.26 (s, 3 H), 1.21 (s, 3H), 1.23 (s, 9H), 1.19 (s, 9H), 0.84 (d, J ) 6.5 Hz, 3H); 13C NMR (50 MHz, some peaks in the aliphatic region overlap) δ 197.3, 177.8, 177.0, 165.0, 151.3, 144.4, 137.3, 128.6, 128.2, 128.1, 127.9, 125.4, 125.1, 122.1, 75.3, 74.7, 72.6, 71.8, 69.1, 50.4, 41.6, 39.9, 38.8, 34.5, 34.1, 31.2, 27.1, 26.6, 25.9, 21.7. 24b: 1H NMR (400 MHz, selected data assigned from a mixture (95:5) of diastereomers 23b and 24b) δ 6.13 (dd, J ) 15.8, 4.5 Hz, 1H), 5.28 (dd, J ) 15.8, 1.7 Hz, 1H). (Z)-(4S,6S,8R)-6-(Benzyloxy)-4,8-bis((2,2-dimethylpropionyl)oxy)-9-oxonon-2-enoic Acid (1R,2S,5R)-5-Methyl2-(1-methyl-1-phenylethyl)cyclohexyl Ester (25b).46 Prepared from 3d and 14b in 62% yield; (Z):(E) ) 98:2, diastereomeric ratio 25b:26b ) 98:2. 25b: 1H NMR (250 MHz, selected data) δ 9.48 (s, 1H), 7.38-7.16 (m, 9H), 7.13-7.04 (m, 1H), 6.24 (dddd, J ) 9.3, 7.5, 3.3, 1.3 Hz, 1H), 5.79 (dd, J ) 11.6, 7.4 Hz, 1H), 5.20 (dd, J ) 10.2, 3.4 Hz, 1H), 5.04 (dd, J ) 11.6, 1.4 Hz, 1H), 4.82 (ddd [app td], J ) 10.7, 4.3 Hz, 1H), 4.60 (d, J ) 11.1 Hz, 1H), 4.40 (d J ) 11.1 Hz, 1H), 3.753.65 (m, 1H), 1.21 (s, 9H), 1.14 (s, 9H), 0.86 (d, J ) 6.5 Hz, 3H); 13C NMR (125 MHz, some peaks in the aliphatic region overlap) δ 197.6, 177.7, 177.4, 164.4, 151.5, 146.7, 137.9, 128.4, 128.1, 127.9, 127.7, 125.4, 125.0, 120.9, 75.5, 74.4, 72.3, 71.7, 69.8, 50.5, 41.7, 39.7, 38.8, 34.6, 31.3, 27.7, 27.1, 26.6, 25.2, 21.8. A sample enriched in diastereomer 26b was obtained from a different experiment. 26b: 1H NMR (250 MHz, selected data) δ 9.46 (s, 1 H), 5.82 (dd, J ) 11.5, 7.5 Hz, 1H.

Acknowledgment. We gratefully acknowledge financial support from the Danish Technical Science Research Council (STVF), the Swedish Natural Science Research Council (NFR), the Carl Trygger Foundation (Sweden), the Pharmacia Research Foundation (Sweden), the Foundation Bengt Lundqvists Minne (Sweden), the Helge Ax:son Johnson Foundation (Sweden), the Technical University of Denmark, and the Royal Institute of Technology. J.S.T. is grateful to Pfizer Central Research (USA) for an Undergraduate Research Fellowship and to the National Science Foundation (USA) for a Graduate Fellowship. NMR instrumentation at the Royal Institute of Technology was funded by the Wallenberg Foundation. Professor Toshiro Harada, Kyoto Institute of Technology, is acknowledged for providing additional experimental information concerning the preparation of diol 5. We also thank Ms Gurli Hammarberg for assistance with spectroscopic analyses. Supporting Information Available: Additional experimental and characterization data (31 pages). This material is contained in libraries on microfiche, immediately follows this article in the microfilm version of the journal, and can be ordered from the ACS; see any current masthead page for ordering information. JO981102Z